Extended high-frequency hearing and head orientation cues benefit children during speech-in-speech recognition

Gender and speech material effects on the long-term average speech spectrum, including at extended high frequencies

Gender and language effects on the long-term average speech spectrum (LTASS) have been reported, but typically using recordings that were bandlimited and/or failed to accurately capture extended high frequencies (EHFs). Accurate characterization of the full-band LTASS is warranted given recent data on the contribution of EHFs to speech perception. The present study characterized the LTASS for high-fidelity, anechoic recordings of males and females producing Bamford-Kowal-Bench sentences, digits, and unscripted narratives. Gender had an effect on spectral levels at both ends of the spectrum: males had higher levels than females below approximately 160 Hz, owing to lower fundamental frequencies; females had ∼4 dB higher levels at EHFs, but this effect was dependent on speech material. Gender differences were also observed at ∼300 Hz, and between 800 and 1000 Hz, as previously reported. Despite differences in phonetic content, there were only small, gender-dependent differences in EHF levels across speech materials. EHF levels were highly correlated across materials, indicating relative consistency within talkers. Our findings suggest that LTASS levels at EHFs are influenced primarily by talker and gender, highlighting the need for future research to assess whether EHF cues are more audible for female speech than for male speech.

Factors influencing the minimum audible change in talker head orientation cues using diotic stimuli

The perception of a talker's head orientation is an ecologically relevant task. Humans are able to discriminate changes in talker head orientation using acoustic cues. Factors that may influence measures of this ability have not been well characterized. Here, we examined the minimum audible change in head orientation cues (MACHO) using diotic stimuli. The effects of several factors were tested: talker and gender, stimulus bandwidth (full-band vs low-pass filtered at 8 or 10 kHz), transducer (loudspeaker vs headphone), stimulus uncertainty (interleaved vs blocked presentation of four talkers), and vocal production mode (speech vs singing). The best performance of ∼41° was achieved for full-band, blocked presentation of speech over a loudspeaker. Greater stimulus uncertainty (interleaved presentation) worsened the MACHO by 26%. Bandlimiting at 8 and 10 kHz worsened performance by an additional 22% and 14%, respectively. At equivalent overall sound levels, performance was better for speech than for singing. There was some limited evidence for the transducer influencing the MACHO. These findings suggest the MACHO relies on multiple factors manipulated here. One of the largest, consistent effects was that of talker, suggesting head orientation cues are highly dependent on individual talker characteristics. This may be due to individual variability in speech directivity patterns.

Band importance for speech-in-speech recognition in the presence of extended high-frequency cues

Band importance functions for speech-in-noise recognition, typically determined in the presence of steady background noise, indicate a negligible role for extended high frequencies (EHFs; 8-20 kHz). However, recent findings indicate that EHF cues support speech recognition in multi-talker environments, particularly when the masker has reduced EHF levels relative to the target. This scenario can occur in natural auditory scenes when the target talker is facing the listener, but the maskers are not. In this study, we measured the importance of five bands from 40 to 20 000 Hz for speech-in-speech recognition by notch-filtering the bands individually. Stimuli consisted of a female target talker recorded from 0° and a spatially co-located two-talker female masker recorded either from 0° or 56.25°, simulating a masker either facing the listener or facing away, respectively. Results indicated peak band importance in the 0.4-1.3 kHz band and a negligible effect of removing the EHF band in the facing-masker condition. However, in the non-facing condition, the peak was broader and EHF importance was higher and comparable to that of the 3.3-8.3 kHz band in the facing-masker condition. These findings suggest that EHFs contain important cues for speech recognition in listening conditions with mismatched talker head orientations.

Developmental Effects in the "Vocale Rapide dans le Bruit" Speech-in-Noise Identification Test: Reference Performances of Normal-Hearing Children and Adolescents

PURPOSE

The main objective of this study was to assess the existence of developmental effects on the performance of the Vocale Rapide dans le Bruit (VRB) speech-in-noise (SIN) identification test that was recently developed for the French language and to collect reference scores for children and adolescents.

METHOD

Seventy-two native French speakers, aged 10-20 years, participated in the study. Each participant listened and repeated four lists of eight sentences, each containing three key words to be scored. The sentences were presented in free field at different signal-to-noise ratios (SNRs) using a four-talker babble noise. The SNR yielding 50% of correct repetitions of key words (SNR50) was recorded for each list.

RESULTS

A strong relationship between age and SNR50 was found, better performance occurring with increasing age (average drop in SNR50 per year: 0.34 dB). Large differences (Cohen's d ≥ 1.2) were observed between the SNR50 achieved by 10- to 13-year-old participants and those of adults. For participants aged 14-15 years, the difference fell just above the 5% level of significance. No effects of hearing thresholds or level of education were observed.

CONCLUSIONS

The study confirms the existence of developmental effects on SIN identification performance as measured using the VRB test and provides reference data for taking into account these effects during clinical practice. Explanations as to why age effects perdure during adolescence are discussed.

Effects of Stimulus Type on 16-kHz Detection Thresholds

OBJECTIVES

Audiometric testing typically does not include frequencies above 8 kHz. However, recent research suggests that extended high-frequency (EHF) sensitivity could affect hearing in natural communication environments. Clinical assessment of hearing often employs pure tones and frequency-modulated (FM) tones interchangeably regardless of frequency. The present study was designed to evaluate how the stimulus chosen to measure EHF thresholds affects estimates of hearing sensitivity.

DESIGN

The first experiment used standard audiometric procedures to measure 8- and 16-kHz thresholds for 5- to 28-year olds with normal hearing in the standard audiometric range (250 to 8000 Hz). Stimuli were steady tones, pulsed tones, and FM tones. The second experiment tested 18- to 28-year olds with normal hearing in the standard audiometric range using psychophysical procedures to evaluate how changes in sensitivity as a function of frequency affect detection of stimuli that differ with respect to bandwidth, including bands of noise. Thresholds were measured using steady tones, pulsed tones, FM tones, narrow bands of noise, and one-third-octave bands of noise at a range of center frequencies in one ear.

RESULTS

In experiment 1, thresholds improved with increasing age at 8 kHz and worsened with increasing age at 16 kHz. Thresholds for individual participants were relatively similar for steady, pulsed, and FM tones at 8 kHz. At 16 kHz, mean thresholds were approximately 5 dB lower for FM tones than for steady or pulsed tones. This stimulus effect did not differ as a function of age. Experiment 2 replicated this greater stimulus effect at 16 kHz than at 8 kHz and showed that the slope of the audibility curve accounted for these effects.

CONCLUSIONS

Contrary to prior expectations, there was no evidence that the choice of stimulus type affected school-age children more than adults. For individual participants, audiometric thresholds at 16 kHz were as much as 20 dB lower for FM tones than for steady tones. Threshold differences across stimuli at 16 kHz were predicted by differences in audibility across frequency, which can vary markedly between listeners. These results highlight the importance of considering spectral width of the stimulus used to evaluate EHF thresholds.

The effects of face masks on speech-in-speech recognition for children and adults

OBJECTIVES

This study explored the effects of different face masks on school-age children's and young adults' word recognition.

DESIGN

Speech recognition thresholds were measured adaptively in a two-talker speech masker using a closed-set picture pointing task. Target words were recorded by a female talker in five conditions: no mask, transparent mask, face shield, N95 mask and surgical mask.

STUDY SAMPLES

Thirty children (8-12 years) and 25 adults (18-25 years) with normal hearing.

RESULTS

Both children's and adults' word recognition was most negatively impacted by the face shield. Children's recognition was also impaired by the transparent mask. No negative effects were observed for the N95 or surgical mask for either age group.

CONCLUSION

School-age children, like young adults, are negatively affected by face masks when recognising speech in a two-talker speech masker, but the effects depend on the type of face mask being worn. Acoustic analyses suggest that the reflective materials used for masks impact speech signal quality and impair word recognition.

Classification of indexical and segmental features of human speech using low- and high-frequency energya)

The high-frequency region (above 4-5 kHz) of the speech spectrum has received substantial research attention over the previous decade, with a host of studies documenting the presence of important and useful information in this region. The purpose of the current experiment was to compare the presence of indexical and segmental information in the low- and high-frequency region of speech (below and above 4 kHz) and to determine the extent to which information from these regions can be used in a machine learning framework to correctly classify indexical and segmental aspects of the speech signal. Naturally produced vowel segments produced by ten male and ten female talkers were used as input to a temporal dictionary ensemble classification model in unfiltered, low-pass filtered (below 4 kHz), and high-pass filtered (above 4 kHz) conditions. Classification performance in the unfiltered and low-pass filtered conditions was approximately 90% or better for vowel categorization, talker sex, and individual talker identity tasks. Classification performance for high-pass filtered signals composed of energy above 4 kHz was well above chance for the same tasks. For several classification tasks (i.e., talker sex and talker identity), high-pass filtering had minimal effect on classification performance, suggesting the preservation of indexical information above 4 kHz.

Explainable machine learning reveals the relationship between hearing thresholds and speech-in-noise recognition in listeners with normal audiograms

Some individuals complain of listening-in-noise difficulty despite having a normal audiogram. In this study, machine learning is applied to examine the extent to which hearing thresholds can predict speech-in-noise recognition among normal-hearing individuals. The specific goals were to (1) compare the performance of one standard (GAM, generalized additive model) and four machine learning models (ANN, artificial neural network; DNN, deep neural network; RF, random forest; XGBoost; eXtreme gradient boosting), and (2) examine the relative contribution of individual audiometric frequencies and demographic variables in predicting speech-in-noise recognition. Archival data included thresholds (0.25-16 kHz) and speech recognition thresholds (SRTs) from listeners with clinically normal audiograms (n = 764 participants or 1528 ears; age, 4-38 years old). Among the machine learning models, XGBoost performed significantly better than other methods (mean absolute error; MAE = 1.62 dB). ANN and RF yielded similar performances (MAE = 1.68 and 1.67 dB, respectively), whereas, surprisingly, DNN showed relatively poorer performance (MAE = 1.94 dB). The MAE for GAM was 1.61 dB. SHapley Additive exPlanations revealed that age, thresholds at 16 kHz, 12.5 kHz, etc., on the order of importance, contributed to SRT. These results suggest the importance of hearing in the extended high frequencies for predicting speech-in-noise recognition in listeners with normal audiograms.

Speech understanding and extended high-frequency hearing sensitivity in blast-exposed veteransa)

Auditory difficulties reported by normal-hearing Veterans with a history of blast exposure are primarily thought to stem from processing deficits in the central nervous system. However, previous work on speech understanding in noise difficulties in this patient population have only considered peripheral hearing thresholds in the standard audiometric range. Recent research suggests that variability in extended high-frequency (EHF; >8 kHz) hearing sensitivity may contribute to speech understanding deficits in normal-hearing individuals. Therefore, this work was designed to identify the effects of blast exposure on several common clinical speech understanding measures and EHF hearing sensitivity. This work also aimed to determine whether variability in EHF hearing sensitivity contributes to speech understanding difficulties in normal-hearing blast-exposed Veterans. Data from 41 normal- or near-normal-hearing Veterans with a history of blast exposure and 31 normal- or near-normal-hearing control participants with no history of head injury were employed in this study. Analysis identified an effect of blast exposure on several speech understanding measures but showed no statistically significant differences in EHF thresholds between participant groups. Data showed that variability in EHF hearing sensitivity did not contribute to group-related differences in speech understanding, although study limitations impact interpretation of these results.

Differential benefits of unmasking extended high-frequency content of target or background speech

Current evidence supports the contribution of extended high frequencies (EHFs; >8 kHz) to speech recognition, especially for speech-in-speech scenarios. However, it is unclear whether the benefit of EHFs is due to phonetic information in the EHF band, EHF cues to access phonetic information at lower frequencies, talker segregation cues, or some other mechanism. This study investigated the mechanisms of benefit derived from a mismatch in EHF content between target and masker talkers for speech-in-speech recognition. EHF mismatches were generated using full band (FB) speech and speech low-pass filtered at 8 kHz. Four filtering combinations with independently filtered target and masker speech were used to create two EHF-matched and two EHF-mismatched conditions for one- and two-talker maskers. Performance was best with the FB target and the low-pass masker in both one- and two-talker masker conditions, but the effect was larger for the two-talker masker. No benefit of an EHF mismatch was observed for the low-pass filtered target. A word-by-word analysis indicated higher recognition odds with increasing EHF energy level in the target word. These findings suggest that the audibility of target EHFs provides target phonetic information or target segregation and selective attention cues, but that the audibility of masker EHFs does not confer any segregation benefit.

Predicting Behavioral Threshold at 6 and 8 kHz for Children and Adults Based on the Auditory Brainstem Response

PURPOSE

Common clinical application of auditory brainstem response (ABR) testing is limited to 0.25-4 kHz. Prior research has demonstrated associations between ABR and behavioral thresholds for tone burst stimuli > 4 kHz in adults, but there are no comparable data for children. The ability to predict behavioral thresholds > 4 kHz clinically based on the ABR would provide valuable audiologic information for individuals who are unable to provide behavioral thresholds. This study included children with hearing loss and children with normal hearing to determine the association between ABR and behavioral thresholds at 6 and 8 kHz.

METHOD

ABR and behavioral thresholds were obtained for children ages 4.7-16.7 years (M = 10.5, SD = 3.4) with sensorineural hearing loss (n = 24) or normal hearing sensitivity (n = 16) and for adults ages 18.4-54.4 years (M = 32.7, SD = 10.4) with sensorineural hearing loss (n = 13) or normal hearing sensitivity (n = 11). Thresholds obtained for 6 and 8 kHz using ABR and conventional audiometry were compared.

RESULTS

Differences between ABR and behavioral thresholds averaged 5-6 dB for both children and adults for both test frequencies, with differences of ≤ 20 dB in all instances. Linear mixed modeling for data from participants with hearing loss suggested that ABR threshold is a good predictor of behavioral threshold at 6 and 8 kHz for both children and adults. Test specificity was 100%; no participants with behavioral thresholds ≤ 20 dB HL had ABR thresholds > 25 dB nHL.

CONCLUSIONS

Initial evidence suggests that ABR testing at 6 and 8 kHz is reliable for estimating behavioral threshold in listeners with hearing loss and accurately identifies normal hearing sensitivity. The results of this study contribute to efforts to improve outcomes for vulnerable populations by reducing barriers to clinical implementation of ABR testing at > 4 kHz.

Brainstem auditory physiology in children with listening difficulties. Hear Res 2023;429:108705. [PMID: 36709582 PMCID: PMC10152893 DOI: 10.1016/j.heares.2023.108705]

Children who have listening difficulties (LiD) despite having normal audiometry are often diagnosed as having an auditory processing disorder. A lack of evidence regarding involvement of specific auditory mechanisms has limited development of effective treatments for these children. Here, we examined electrophysiologic evidence for brainstem pathway mechanisms in children with and without defined LiD. We undertook a prospective controlled study of 132 children aged 6-14 years with normal pure tone audiometry, grouped into LiD (n = 63) or Typically Developing (TD; n = 69) based on scores on the Evaluation of Children's Listening and Processing Skills (ECLiPS), a validated caregiver report. The groups were matched on age at test, sex, race, and ethnicity. Neither group had diagnoses of major neurologic disorder, intellectual disability, or brain injuries. Both groups received a test battery including a measure of receptive speech perception against distractor speech, Listening in Spatialized Noise - Sentences (LiSN-S), along with multiple neurophysiologic measures that tap afferent and efferent auditory subcortical pathways. Group analysis showed that participants with LiD performed significantly poorer on all subtests of the LiSN-S. The LiD group had significantly greater wideband middle ear muscle reflex (MEMR) growth functions in the left ear, and shorter Wave III and Wave V latencies in auditory brainstem responses (ABR). Across individual participants, shorter latency ABR Wave V correlated significantly with poorer parent report of LiD (ECLiPS composite). Greater MEMR growth functions also correlated with poorer ECLiPS scores and reduced LiSN-S talker advantage. The LiD and TD groups had equivalent summating potentials, compound action potentials, envelope-following responses, and binaurally activated medial olivocochlear reflexes. In conclusion, there was no evidence for auditory synaptopathy for LiD. Evidence for brainstem differences in the LiD group was interpreted as increased central gain, with shorter ABR Wave III and V latencies and steeper MEMR growth curves. These differences were related to poorer parent report and speech perception in competing speech ability.

Extending the High-Frequency Bandwidth and Predicting Speech-in-Noise Recognition: Building on the Work of Pat Stelmachowicz

Recent work has demonstrated that high-frequency (>6 kHz) and extended high-frequency (EHF; >8 kHz) hearing is valuable for speech-in-noise recognition. Several studies also indicate that EHF pure-tone thresholds predict speech-in-noise performance. These findings contradict the broadly accepted "speech bandwidth" that has historically been limited to below 8 kHz. This growing body of work is a tribute to the work of Pat Stelmachowicz, whose research was instrumental in revealing the limitations of the prior speech bandwidth work, particularly for female talkers and child listeners. Here, we provide a historical review that demonstrates how the work of Stelmachowicz and her colleagues paved the way for subsequent research to measure effects of extended bandwidths and EHF hearing. We also present a reanalysis of previous data collected in our lab, the results of which suggest that 16-kHz pure-tone thresholds are consistent predictors of speech-in-noise performance, regardless of whether EHF cues are present in the speech signal. Based on the work of Stelmachowicz, her colleagues, and those who have come afterward, we argue that it is time to retire the notion of a limited speech bandwidth for speech perception for both children and adults.

Systematic Review of Literature on Speech Intelligibility and Classroom Acoustics in Elementary Schools

PURPOSE

Good verbal signals and low background noise are key factors for all children to maximize understanding of what is being taught. Classroom shape, surroundings, and even furnishings change how the environment "sounds" and how speech is "heard" in the classroom. Classroom acoustics is perhaps one of the most important, but often least considered, factors when designing a classroom. This systematic review aimed to characterize the relationship between intelligibility of speech and room acoustics in elementary schools based on the available evidence.

METHOD

Eligible studies were identified using two computerized databases: PubMed and Scopus. In total, 23 publications met our inclusion criteria: (a) Participants must have been from elementary schools, (b) acoustic characterization of the classroom must have been provided, (c) intelligibility tests must have been performed, and (d) articles were written in English.

RESULTS

After identifying the parameters and tests used to quantify the intelligibility of speech, the speech intelligibility scores were analyzed in relation with acoustical parameters found in the articles, particularly signal-to-noise ratio and speech transmission index. Our results highlighted the negative effect on intelligibility associated with poor transmission of the speech and poor classroom acoustics caused by long reverberation times and high background noise.

CONCLUSION

Good classroom acoustics is needed to improve speech intelligibility and, therefore, increase children's academic success.

Auditory Brainstem Responses at 6 and 8 kHz in Infants With Normal Hearing

PURPOSE

Normative auditory brainstem response (ABR) data for infants and young children are available for 0.25-4 kHz, limiting clinical assessment to this range. As such, the high-frequency hearing sensitivity of infants and young children remains unknown until behavioral testing can be completed, often not until late preschool or early school ages. The purpose of this study was to obtain normative ABR data at 6 and 8 kHz in young infants.

METHOD

Participants were 173 full-term infants seen clinically for ABR testing at 0.4-6.7 months chronological age (M = 1.4 months, SD = 1.0), 97% of whom were ≤ 12 weeks chronological age. Stimuli included 6 and 8 kHz tone bursts presented at a rate of 27.7/s or 30.7/s using Blackman window gating with six cycles (6 kHz) or eight cycles (8 kHz) rise/fall time and no plateau. Presentation levels included 20, 40, and 60 dB nHL. The ABR threshold was estimated in 5- to 10-dB steps.

RESULTS

As previously observed with lower frequency stimuli, ABR waveforms obtained in response to 6 and 8 kHz tone bursts decreased in latency with increasing intensity and increasing age. Latency was shorter for 8-kHz tone bursts than 6-kHz tone bursts. Data tables are presented for clinical reference for infants ≤ 4 weeks, 4.1-8 weeks, and 8.1-12 weeks chronological age including median ABR latency for Waves I, III, and V and the upper and lower boundaries of the 90% prediction interval. Interpeak Latencies I-III, III-V, and I-V are also reported.

CONCLUSION

The results from this study demonstrate that ABR assessment at 6 and 8 kHz is feasible for young infants within a standard clinical appointment and provide reference data for clinical interpretation of ABR waveforms for frequencies above 4 kHz.

Hearing Impairment in the Extended High Frequencies in Children Despite Clinically Normal Hearing

OBJECTIVES

Pediatric hearing impairment, regardless of degree and type, has a detrimental effect on speech perception, cognition, oral language development, academic outcomes, and literacy. Hearing assessment in the clinic is limited to 8 kHz although humans can hear up to 20 kHz. Hearing impairment in the extended high frequencies (EHFs > 8 kHz) can occur despite clinically normal hearing. However, to date, the nature and effects of EHF hearing impairment in children remain unknown. The goals of the present study were to determine the effects of EHF hearing impairment on speech-in-noise recognition in children and to examine whether hearing impairment in the EHFs is associated with altered cochlear functioning in the standard frequencies.

DESIGN

A volunteer sample of 542 participants (4 to 19 years) with clinically normal audiograms were tested. Participants identified with EHF impairment were assigned as cases in a subsequent case-control study. EHF loss was defined as hearing thresholds greater than 20 dB in at least one EHFs (10, 12.5, or 16 kHz). Speech recognition thresholds in multi-talker babble were measured using the digit triplet test. Distortion product otoacoustic emissions ( f2 = 2, 3, 4, and 5 kHz) were measured to assess cochlear functioning.

RESULTS

Thresholds in the EHFs were as reliable as those in the standard frequency range. Thirty-eight children had EHF hearing impairment regardless of a clinically normal audiogram. A linear mixed-effects model revealed that children with EHF hearing impairment had higher (poorer) mean speech recognition threshold than children with normal EHF sensitivity ( estimate = 2.14 dB, 95% CI: 1.36 to 3.92; effect size = small). The overall magnitude of distortion product otoacoustic emissions was lower for children with EHF impairment ( estimate = -2.47 dB, 95% CI: -4.60 to -0.73; effect size = medium). In addition, the pure-tone average for standard audiometric frequencies was relatively higher for EHF-impaired children ( estimate = 3.68 dB, 95% CI: 2.56 to 4.80; effect size = small).

CONCLUSIONS

Hearing impairment in the EHFs is common in children despite clinically normal hearing and can occur without a history of otitis media. EHF impairment is associated with poorer speech-in-noise recognition and preclinical cochlear deficits in the lower frequencies where hearing thresholds are normal. This study highlights the clinical need to identify EHF impairments in children.

Transcriptome Analyses Provide Insights into the Auditory Function in Trachemys scripta elegans

An auditory ability is essential for communication in vertebrates, and considerable attention has been paid to auditory sensitivity in mammals, birds, and frogs. Turtles were thought to be deaf for a long time; however, recent studies have confirmed the presence of an auditory ability in Trachemys scripta elegans as well as sex-related differences in hearing sensitivity. Earlier studies mainly focused on the morphological and physiological functions of the hearing organ in turtles; thus, the gene expression patterns remain unclear. In this study, 36 transcriptomes from six tissues (inner ear, tympanic membrane, brain, eye, lung, and muscle) were sequenced to explore the gene expression patterns of the hearing system in T. scripta elegans. A weighted gene co-expression network analysis revealed that hub genes related to the inner ear and tympanic membrane are involved in development and signal transduction. Moreover, we identified six differently expressed genes (GABRA1, GABRG2, GABBR2, GNAO1, SLC38A1, and SLC12A5) related to the GABAergic synapse pathway as candidate genes to explain the differences in sexually dimorphic hearing sensitivity. Collectively, this study provides a critical foundation for genetic research on auditory functions in turtles.

On the use of the TIMIT, QuickSIN, NU-6, and other widely used bandlimited speech materials for speech perception experiments

The use of spectrally degraded speech signals deprives listeners of acoustic information that is useful for speech perception. Several popular speech corpora, recorded decades ago, have spectral degradations, including limited extended high-frequency (EHF) (>8 kHz) content. Although frequency content above 8 kHz is often assumed to play little or no role in speech perception, recent research suggests that EHF content in speech can have a significant beneficial impact on speech perception under a wide range of natural listening conditions. This paper provides an analysis of the spectral content of popular speech corpora used for speech perception research to highlight the potential shortcomings of using bandlimited speech materials. Two corpora analyzed here, the TIMIT and NU-6, have substantial low-frequency spectral degradation (<500 Hz) in addition to EHF degradation. We provide an overview of the phenomena potentially missed by using bandlimited speech signals, and the factors to consider when selecting stimuli that are sensitive to these effects.

Reconsidering commonly used stimuli in speech perception experiments

This paper examines some commonly used stimuli in speech perception experiments and raises questions about their use, or about the interpretations of previous results. The takeaway messages are: 1) the Hillenbrand vowels represent a particular dialect rather than a gold standard, and English vowels contain spectral dynamics that have been largely underappreciated, 2) the /ɑ/ context is very common but not clearly superior as a context for testing consonant perception, 3) /ɑ/ is particularly problematic when testing voice-onset-time perception because it introduces strong confounds in the formant transitions, 4) /dɑ/ is grossly overrepresented in neurophysiological studies and yet is insufficient as a generalized proxy for "speech perception," and 5) digit tests and matrix sentences including the coordinate response measure are systematically insensitive to important patterns in speech perception. Each of these stimulus sets and concepts is described with careful attention to their unique value and also cases where they might be misunderstood or over-interpreted.

Effect of Masker Head Orientation, Listener Age, and Extended High-Frequency Sensitivity on Speech Recognition in Spatially Separated Speech

OBJECTIVES

Masked speech recognition is typically assessed as though the target and background talkers are all directly facing the listener. However, background speech in natural environments is often produced by talkers facing other directions, and talker head orientation affects the spectral content of speech, particularly at the extended high frequencies (EHFs; >8 kHz). This study investigated the effect of masker head orientation and listeners' EHF sensitivity on speech-in-speech recognition and spatial release from masking in children and adults.

DESIGN

Participants were 5- to 7-year-olds (n = 15) and adults (n = 34), all with normal hearing up to 8 kHz and a range of EHF hearing thresholds. Speech reception thresholds (SRTs) were measured for target sentences recorded from a microphone directly in front of the talker's mouth and presented from a loudspeaker directly in front of the listener, simulating a target directly in front of and facing the listener. The maskers were two streams of concatenated words recorded from a microphone located at either 0° or 60° azimuth, simulating masker talkers facing the listener or facing away from the listener, respectively. Maskers were presented in one of three spatial conditions: co-located with the target, symmetrically separated on either side of the target (+54° and -54° on the horizontal plane), or asymmetrically separated to the right of the target (both +54° on the horizontal plane).

RESULTS

Performance was poorer for the facing than for the nonfacing masker head orientation. This benefit of the nonfacing masker head orientation, or head orientation release from masking (HORM), was largest under the co-located condition, but it was also observed for the symmetric and asymmetric masker spatial separation conditions. SRTs were positively correlated with the mean 16-kHz threshold across ears in adults for the nonfacing conditions but not for the facing masker conditions. In adults with normal EHF thresholds, the HORM was comparable in magnitude to the benefit of a symmetric spatial separation of the target and maskers. Although children benefited from the nonfacing masker head orientation, their HORM was reduced compared to adults with normal EHF thresholds. Spatial release from masking was comparable across age groups for symmetric masker placement, but it was larger in adults than children for the asymmetric masker.

CONCLUSIONS

Masker head orientation affects speech-in-speech recognition in children and adults, particularly those with normal EHF thresholds. This is important because masker talkers do not all face the listener under most natural listening conditions, and assuming a midline orientation would tend to overestimate the effect of spatial separation. The benefits associated with EHF audibility for speech-in-speech recognition may warrant clinical evaluation of thresholds above 8 kHz.